This invention relates to refractories for use in the production of molten aluminum and aluminum alloys, and more specifically, it concerns a refractory containing minor additions of a water-insoluble, chlorine-containing material such as polyvinyl chloride (PVC) powder.
The containment of molten aluminum presents a difficult challenge to the refractory specialist. Molten aluminum, especially superheated aluminum, is extremely fluid and corrosive. Common alloying agents, such as magnesium and lithium, enhance the reactive and corrosive power of the alloy. Refractories used to contain aluminum tend to have open porosity. Once molten aluminum penetrates into the refractory, chemical reactions occur between the aluminum, its alloying agents, and the silica-containing phases (or other reducible phases) in the refractory. The products of these reactions, namely silicon and alumina, cause severe problems. Silicon contaminates molten aluminum thereby decreasing its commercial value. Alumina forms a dross buildup which decreases the holding capacity of the furnace and makes heat transfer from the flame to the bath more difficult. In addition, the reactions between the aluminum, its alloying agents and the refractory disrupt the refractory and, thereby, decrease the life of the furnace.
There has been an on-going effort in the refractories industry to provide a more aluminum-resistant refractory. One approach dealt with aluminum orthophosphate bonding of alumino-silicate refractories as a means to decrease molten aluminum penetration. While this type of bonding has provided some improvement, the need exists for a less reactive refractory to contain aluminum. This need has arisen due to the increasingly tighter specifications on silicon and phosphorous levels in the final product. Because of this concern, some aluminum producers have recently excluded the use of phosphate bonded compositions in certain wear-prone areas of aluminum reverberatory furnaces such as sidewalls and bellybands.
Another approach has been the addition of borosilicate frits to calcium aluminate cement-bonded refractories. A disadvantage of the addition of borosilicate frits is that these additions are slightly soluble in water and they have an acidic pH. Their use in low cement castables sometimes prevents the refractory from hardening at room temperature. A further disadvantage of these frits is their lack of refractoriness at elevated temperatures. This causes loss of strength which is a concern since at times aluminum-producing furnaces often experience temperature excursions 1.7 times greater than that needed to melt aluminum.
U.S. Pat. No. 4,806,509 issued to Porterfield on Feb. 21, 1989, discloses yet another approach to providing an aluminum resistant refractory composition. In particular, Porterfield discloses a refractory including a refractory aggregate, a binder, and a synthetic fused aluminum borate additive for imparting aluminum resistance. This additive is a calcium fluoride aluminum borate aggregate with the calcium fluoride dispersed within an insoluble structure provided by the fused aluminum borate. The synthetic additive of Porterfield is produced by blending boric acid, calcined alumina and fluorspor. Although the synthetic additive of Porterfield may provide the refractory with improved resistance to aluminum, at high addition levels the synthetic additive may be detrimental to the physical properties of the refractory and increase the costs of the product containing such a synthetic additive.
U.S. Pat. No. 4,510,253 issued to Felice et al on Apr. 9, 1985, discloses a lightweight refractory composition for use in contact with molten aluminum alloys. This lightweight refractory includes ceramic fibers, binders, and an additive containing 9Al.sub.2 O.sub.3.2-B.sub.2 O.sub.3. This additive is preferably aluminum boron slag which is a byproduct of the production of ferro-boron. The aluminum boron slag utilized by Felice et al may contain impurities which decrease the refractoriness of products which contain this additive. The slag may also contain compounds which retard set times and destroy hydraulic bonding.
In light of the foregoing, the need clearly exists for a refractory that contains an aluminum penetration inhibitor which is compatible in acid-bonded systems, cement-bonded systems, is not water soluble, does not adversely affect hot strength, and is relatively inexpensive to produce.